High connotation phenotype analysis is increasingly being applied in many research fields, including gene function research, drug development, and toxicology. The advantage of this method is that it objectively obtains multidimensional information from a single level, can analyze and describe the state of individual cells, and compare it with the overall analysis data. However, many traditional phenotype analyses subjectively select only a few important indicators to analyze in the pathway to be studied. Therefore, in traditional phenotype analysis, most of the biologically related changes caused by experimental processing are easily overlooked. The phenotype analysis of a cell is composed of thousands of quantifiable indicators that characterize its state. These indicators contain information extracted from the expression, structure, and distribution of biomarkers. Usually, these markers are recognized by organelles and provide information about their structural and spatial relationships with other subcellular structures. The classification information of these cells can provide customers with effective analytical basis for studying compounds, small molecule substances, or gene disturbances. Compounds with similar mechanisms of action (MOA) often lead to similar cellular morphological changes. Therefore, the comparison between phenotype spectra can provide a basis for the MOA of new compounds. Similarly, genetic interference in the same pathway often leads to similar phenotype profiles, indicating that phenotype profiles can be used for high-throughput functional genomics research.